This project will develop a framework based on mathematical modeling that a) describes the mechanism by which synaptic plasticity emerges from molecular processes regulating gene transcription, and b) tests mechanistic hypotheses, such as proposed roles of specific protein kinases. The project builds upon our previous model describing aspects of the gene and protein network responsible for long-term synaptic facilitation (LTF) and the formation of long-term memory (LTM) in the mollusk Aplysia. This model is based on transcriptional regulation by Ca2+/cAMP response element - binding protein (CREB, termed ApCREBI in Aplysia) and related transcription factors. We will extend this model to incorporate additional elements of gene regulation recently demonstrated to be essential for LTF. In addition, we will develop an analogous model to simulate biochemical events underlying the induction of late long-term synaptic potentiation (L-LTP) in vertebrates. Both LTF and L-LTP are thought to play essential roles in the formation of LTM, and LTF induction and L-LTP induction exhibit mechanistic similarities, such as dependence on MAP kinase activation. Therefore, a modeling framework that can simulate aspects of both LTF and L-LTP induction is likely to significantly increase the understanding of learning mechanisms. The LTF model variant will incorporate additional transcriptional regulators essential for LTF, such as ApCREB2, and ApC/EBP. Bifurcation analysis and pre-programmed integrations will identify key control parameters which are plausible sites of physiological regulation and which, when varied, have important effects on the dynamics of the model. We will then use the model to simulate the results of experimental protocols in which alterations are made in the activity of the transcriptional regulators listed above. A minimal set of variations in key control parameters will be identified that allows simulation of data from these protocols. This approach is likely to help identify the key mechanisms that determine the amount of LTF induced by different training protocols. The L-LTP model variant will be used to simulate three common stimulus protocols that induce hippocampal L-LTP. These protocols are high-frequency (tetanic) stimulation, theta-burst stimulation, and stimulation by forskolin. Parameters will be optimized to fit experimental time courses of nuclear [Ca2+] and of kinase and transcription factor activities. The model will then be used to test the hypothesis that CREB kinases other than protein kinase A, such as ribosomal S6 kinase 2, are primarily responsible for CREB phosphorylation and LTP induction. Preliminary model development and simulations predict that L-LTP induction by a low-frequency burst stimulus protocol does not depend on nuclear CaM kinase activation and consequent CREB phosphorylation.